README.md

Cloud object detection application

This code sample shows you an implementation of object detection based on Azure Cognitive Services.

To run the application:

1. Start chariott runtime by executing cargo run from the root directory

2. Start it by executing cargo run from the examples/applications/cloud-object-detection directory while specifying COGNITIVE_ENDPOINT (i.e. myendpoint.cognitiveservices.azure.com) and COGNITIVE_KEY environment variables.

3. In the root directory create a detect_image.json file with the following message structure:

    {
        "intent": {
            "invoke": {
                "command": "detect",
                "args": [
                    {
                        "any": {
                            "@type": "examples.detection.v1.DetectRequest",
                            "blob": {
                                "media_type": "image/jpg",
                                "bytes": "base64 encoding of the image"
                            }
                        }
                    }
                ]
            }
        },
        "namespace": "sdv.detection"
    }

4. Execute detection with grpcurl -v -plaintext -import-path proto/ \ -import-path examples/applications/proto -use-reflection -proto \ examples/applications/proto/examples/detection/v1/detection.proto -d @ \ localhost:4243 chariott.runtime.v1.ChariottService/Fulfill < \ detect_image.json